BMC Musculoskeletal Disorders BioMed Central · visual estimation of ROMs performed by residents [17]. By comparing values given by the POS to those of the res-ident, our purpose

BioMed CentralBMC Musculoskeletal Disorders

ss
Open AcceResearch articleIs visual estimation of passive range of motion in the pediatric lower limb valid and reliableRami Rachkidi*, Ismat Ghanem, Ibrahim Kalouche, Samer El Hage, Fernand Dagher and Khalil Kharrat
Address: Department of orthopaedic surgery, Hôtel-Dieu de France Hospital, Saint Joseph University. Beyrouth, Lebanon

Email: Rami Rachkidi* - [email protected]; Ismat Ghanem - [email protected]; Ibrahim Kalouche - [email protected]; Samer El Hage - [email protected]; Fernand Dagher - [email protected]; Khalil Kharrat - [email protected]

* Corresponding author

AbstractBackground: Visual estimation (VE) is an essential tool for evaluation of range of motion. Fewpapers discussed its validity in children orthopedics' practice. The purpose of our study was toassess validity and reliability of VE for passive range of motions (PROMs) of children's lower limbs.

Methods: Fifty typically developing children (100 lower limbs) were examined. Visual estimationsfor PROMs of hip (flexion, adduction, abduction, internal and external rotations), knee (flexion andpopliteal angle) and ankle (dorsiflexion and plantarflexion) were made by a pediatric orthopaedicsurgeon (POS) and a 5th year resident in orthopaedics. A last year medical student did goniometricmeasurements. Three weeks later, same measurements were performed to assess reliability ofvisual estimation for each examiner.

Results: Visual estimations of the POS were highly reliable for hip flexion, hip rotations andpopliteal angle (ρc ≥ 0.8). Reliability was good for hip abduction, knee flexion, ankle dorsiflexion andplantarflexion (ρc ≥ 0.7) but poor for hip adduction (ρc = 0.5). Reproducibility for all PROMs wasverified. Resident's VE showed high reliability (ρc ≥ 0.8) for hip flexion and popliteal angle. Goodcorrelation was found for hip rotations and knee flexion (ρc ≥ 0.7). Poor results were obtained forankle PROMs (ρc < 0.6) as well as hip adduction and abduction, the results of which not beingreproducible. Influence of experience was clearly demonstrated for PROMs of hip rotations,adduction and abduction as well as ankle plantarflexion.

Conclusion: Accuracy of VE of passive hip flexion and knee PROMs is high regardless of theexaminer's experience. Same accuracy can be found for hip rotations and abduction whenever VEis performed by an experienced examiner. Goniometric evaluation is recommended for passive hipadduction and for ankle PROMs.

BackgroundPassive range of motion (PROM) measurement is anessential tool for pediatric orthopedists to document dis-ease progression, to plan treatment and to evaluate its

results. While many methods are available for ROM eval-uation, most physicians consider goniometry as the goldstandard and its validity is currently widely accepted espe-cially when measurements are taken by the same exam-

Published: 12 October 2009

BMC Musculoskeletal Disorders 2009, 10:126 doi:10.1186/1471-2474-10-126

Received: 21 October 2008Accepted: 12 October 2009

This article is available from: http://www.biomedcentral.com/1471-2474/10/126

© 2009 Rachkidi et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 1 of 10(page number not for citation purposes)

http://www.biomedcentral.com/1471-2474/10/126

http://creativecommons.org/licenses/by/2.0

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Abstract&list_uids=19822011

http://www.biomedcentral.com/

http://www.biomedcentral.com/info/about/charter/

BMC Musculoskeletal Disorders 2009, 10:126 http://www.biomedcentral.com/1471-2474/10/126

iner within the same session and on the same day [1-16].Passive ROM measures are generally thought to be morereliable in individuals who have normal tone than in indi-viduals with hypertonicity [11-15]. Kilgour et al [15]found that sagittal plane ROM goniometric measures havesimilar levels of reliability in children who have spasticdiplegia and in control children, both within and betweensessions. Validity and reliability of goniometric measure-ments were well studied in the paper of Gajdosik [16],concluding that the universal full-circle goniometer is thepreferred instrument for measuring ROM.

However, in regular daily practice, the majority of clini-cians rely on visual estimations (VE) of angles, the relia-bility of which is yet to be established. In a wide search ofthe current literature, we found few studies assessing thevalidity of this method in evaluating the PROM of one ortwo joints of the lower limb of normal subjects [6,9].

We studied PROM of hips, knees and ankles in a healthypediatric population.

The aim of this study is threefold:

- To assess the validity of visual estimation

- To study the reproducibility of the results (intra-tester reliability)

- To evaluate the influence of experience (inter-testerreliability)

MethodsSubjectsFifty typically developing children (100 lower limbs)without any orthopedic history were examined. Therewere 32 girls and 18 boys with an average age of 8 years (3y-15 y).

ExaminersThe study received ethical approval from our universityEthics Committee and informed consent was given byparents. Three examiners were involved in the study. VEmeasurements were made by a pediatric orthopaedic sur-geon (POS) and a 5th year resident in orthopaedics. A lastyear medical student did the goniometric measurements.

The specialist is an experienced pediatric orthopedic sur-geon (15 years of experience) performing regularly visualestimates for PROMs. The second examiner was a last yearresident in orthopedic surgery (5th year of residency), alsoperforming visual estimation in a regular manner. At leastone study reporting angular values for joint mobility usesvisual estimation of ROMs performed by residents [17].

By comparing values given by the POS to those of the res-ident, our purpose was to assess the influence of experi-ence. The third examiner who did goniometricmeasurements was a last year medical student (7th year ofmedical studies), familiarized with the technique afterthorough explanations given by the POS and a practicesession. Fish and Wingate [18] have stated that even inex-perienced persons can use correctly the goniometer if theyknow the technique.

InstrumentationThe measurement tool was the classical two-arm goniom-eter with a scale marked in 2-degree-increments (figure 1).Validity of the goniometer was verified by measuring tenrandomly selected angles drawn by a computer. For visualestimation, PROM measurements were given by multiplesof 5 degrees.

ProcedureThe study was divided into two parts.

PART 1All relevant bony landmarks on the lower limbs wereexposed and marked on each child (linea alba, middle ofinguinal crease as a marker of the center of the hip, greatertrochanter, longitudinal axis of the patella, lateral femoralcondyle, tibial crest and medial malleolus). We evaluatedall PROM measurements in the supine position except forinternal and external rotations of the hip (prone posi-tion).

Each examiner recorded measurements independently.All motions were passively undertaken and maintained bythe pediatric orthopaedic surgeon (POS) at their maxi-

Two-arm goniometer with a scale marked in 2-degree-incre-mentsFigure 1Two-arm goniometer with a scale marked in 2-degree-increments.



mum magnitude, in such a way as to allow visual estima-tion to be performed for the same amplitude by himselfand the resident from the same visual angle, the residentstanding right behind the surgeon. Then, and while main-taining the same position, the third examiner did thegoniometric measurement to which "visual" examinerswere blinded. Standard techniques of goniometric meas-urements were used with some adjustments to reflect clin-ical practice [19,20].

For the nine motions described below, the three examin-ers stood in such a way as to have their visual angle themost perpendicular possible to the plane of motion.

- Hip flexion (HFX): The POS flexed the thigh over theabdomen. The student placed one of the goniometer'sarms on the examination table and the other along theaxis joining the greater trochanter to the lateral femo-ral condyle (figure 2).

- Hip adduction (HAD): The surgeon adducted thelimb over the other one (adduction associated to somedegree of flexion). The center of the goniometer wasplaced over the middle of the inguinal crease, with oneof its arms parallel to the linea alba and the otherpointing towards the longitudinal axis of the patella(figure 3).

- Hip abduction (HAB): The limb was abducted andthe goniometric measurement was performed asdescribed for hip adduction (figure 4).

- Hip internal rotation (HIR): This motion was done inthe prone position, with 90 degrees of knee flexion.

One of the arms of the goniometer was placed alongthe examination table and the other in line with thetibial crest (figure 5).

- Hip external rotation (HER): same measurement asfor internal rotation but when performing externalrotation of the hip (figure 6).

- Knee flexion (KFX): the center of the goniometer wasplaced over the lateral condyle with one of its armspointing towards the greater trochanter and the otherparallel to the anterior aspect of the leg (figure 7).

- Popliteal angle (PA): Goniometric measurement wasperformed in the same manner as for knee flexion.We've studied this motion instead of knee extension

Hip flexionFigure 2Hip flexion.

Hip adductionFigure 3Hip adduction.

Hip abductionFigure 4Hip abduction.



because the values are more variable and the resultsare more interesting (figure 8).

- Ankle dorsiflexion (ADF): the foot was dorsiflexedtowards the leg. The center of the goniometer wasplaced over the medial malleolus with one of its armsparallel to the anterior aspect of the leg and the otherparallel to the line joining the plantar aspects of theheel and the head of the first metatarsal. We've chosento examine ROMs of ankle as we usually do in a regu-lar clinical setting which is from the medial side on apatient supine with lower limb externally rotated.Knee was in flexion (figure 9).

Hip internal rotationFigure 5Hip internal rotation.

Hip external rotationFigure 6Hip external rotation.

Knee flexionFigure 7Knee flexion.

Popliteal angleFigure 8Popliteal angle.



- Ankle plantarflexion (APF): same goniometric meas-urement as for ankle dorsiflexion but when the foot isplantarflexed (figure 10).

PART 2Three weeks later, the same measurements were per-formed again by the same examiners, on the same chil-dren, at the same place, in the same manner and using thesame goniometer. Our main objective in this part was tostudy the reliability of the VE of each examiner. Also, andby studying independently the results of this part, theoverall number of measurements is doubled. This part ofthe study was done three weeks later. Skeletal landmarkswere identified again. Some cutaneous markings were stillvisible from the first part but the majority was erased. We

tried to reproduce the same markings by adopting thesame techniques of identification of bony landmarks.

Data analysisThe three methods of measurements (two visual estima-tions and one goniometric measurement) yield continu-ous variables. We considered goniometric measurementsas the reference to which corresponding visual estima-tions were compared.

Statistical analysis was done using the test of LIN for cal-culating the concordance correlation coefficient [21,22]and the graphs of Bland and Altman for measuring the dif-ference between VE and goniometric values for eachexaminer [23,24].

The test of Lin [21] calculates a concordance correlationcoefficient (ccc/rho_c/ρc) for agreement on a continuousmeasurement obtained by two persons or methods. Itevaluates the degree to which pairs of observations fall onthe 45° line through the origin. This coefficient rangesfrom zero (no agreement) to one (perfect agreement). The"ccc" is the product of Pearson correlation coefficient (r)by a bias correction factor (Cb)

where

• (ρc) is the Lin concordance correlation coefficient; 0≤ ρc ≤ 1

• (r) is the Pearson correlation coefficient which meas-ures how far each observation deviates from the best-fit line, and is a measure of precision; 0 ≤ r ≤ 1

• (C_b) is a bias correction factor that measures howfar the best-fit line deviates from the 45° line throughthe origin, and is a measure of accuracy; Cb < 1

There are no categorized levels for concordance coefficientvalues but for descriptive reasons we have arbitrarily cho-sen four categories for correlation: High (ρc ≥ 0.8), good(0.7 ≤ ρc < 0.8), fair (0.6 ≤ ρc < 0.7) and poor (ρc < 0.6).

We show in figure 11 an example of graphical representa-tion of correlation between visual estimation of poplitealangle done by the resident and goniometric measurementat the first part of study.

Thus, Lin coefficient combines measures of precision andaccuracy. Pearson correlation coefficient is inappropriatefor measuring agreement between two methods because itestimates the degree of linear association while ignoringsystematic bias. Two methods may have a strong linear

ρc br C= ( ).( )

Ankle dorsiflexionFigure 9Ankle dorsiflexion.

Ankle plantarflexionFigure 10Ankle plantarflexion.



association but a poor agreement [23]. Validation ofreproducibility was done using the confidence intervalapproach [25].

Graphs of Bland and Altman are used to graphically rep-resent results obtained by two methods of measurementand are useful to estimate and represent graphically meas-urement errors. One point is affected for each measure,with the mean value of the two measurements as "x" andtheir difference as "y". A horizontal line is obtained repre-senting the overall mean difference with two other linesrepresenting limits of agreement at 95% (which meansthat 95% of the results are between the limits). The overallmean difference line may be over (overestimation) orunder (underestimation) the zero line.

We show in figure 12 a graphical representation of special-ist's estimations for hip flexion compared to goniometricmeasurements, according to Bland-Altman method.

We reached the three main objectives of our study by:

- Assessing the concordance correlation coefficient[21] and the graphs of Bland-Altman [23] (study ofvalidity of visual estimation)

- Comparing respective values of "ccc" obtained in thetwo parts of study for each examiner seperately byusing the confidence interval method [25] (intra-raterreliability). We've deliberately chosen this method ofevaluation of intra-rater reliability to avoid biasrelated to direct comparison of visual estimations

obtained in parts 1 and 2 of the study. PROMs are verydifficult to reproduce and respective ROMs obtainedin the two parts of the study are certainly different[8,9,15]. This problem is thoroughly explained in thediscussion section.

- Comparing corresponding values of "ccc" for the twoexaminers within each part of the study by using theconfidence interval method [25] (inter-rater reliabilityand influence of experience). This method demon-strates whether a measurement tool is reliable or not,without calculating a coefficient.

Results1. Assessment of validity of visual estimation1.1 Pediatric orthopaedic surgeonCorrelation between visual estimations (VE) of the POSand goniometric measurements for the two parts of studyis summarized in [see Additional file 1 Stable 1]. Wefound high reliability for hip flexion, hip rotations andpopliteal angle (ρc ≥ 0.8). Reliability was good for hipabduction, knee flexion, ankle dorsiflexion and plantar-flexion (ρc ≥ 0.7). Concordance was poor for hip adduc-tion (ρc = 0.5). In all cases, correlation was statisticallysignificant (p < 0.001).

We present in [see Additional file 2 Stable 2] a simplifiedrepresentation of Bland-Altman graphs for the ninePROMs for all measurements of the study (parts 1 and 2)showing mean measurement errors of visual estimation

An example of graphical representation of correlation between visual estimation of popliteal angle (PA) given by the resident (R) and goniometric measurement (G) at the first part of study (1), according to the test of LinFigure 11An example of graphical representation of correla-tion between visual estimation of popliteal angle (PA) given by the resident (R) and goniometric measurement (G) at the first part of study (1), according to the test of Lin. Small red circles: pairs of observation; Red line: best-fit line of observations; Dashed blue line: 45 degrees line; measurements are in degrees.

A representation of Bland-Altman Graph for hip flexion (HFX) estimated by the specialist (S) and compared to gonio-metric measurements (G) in the first part of the study (1)Figure 12A representation of Bland-Altman Graph for hip flex-ion (HFX) estimated by the specialist (S) and com-pared to goniometric measurements (G) in the first part of the study (1). Small red circles: representations for couples of observation with mean value of measurements as "x" and their difference as "y"; Dotted red line: mean meas-urement error; Solid red lines: limits of agreement at 95%; Blue dashed line: zero line; measurements are in degrees.



(in degrees) when compared to goniometric measure-ments. We show 95% confidence interval as well. There isa general tendency to overestimate hip flexion, hip abduc-tion and ankle plantarflexion. Knee PROMs and ankledorsiflexion are underestimated. Correlation was statisti-cally significant in all cases (p < 0.001).

1.2 ResidentThe accuracy of the resident's VE for the two parts of studyis presented in [see Additional file 3 Stable 3]. Hip flexionand popliteal angle showed high reliability (ρc ≥ 0.8). Thecorrelation was good for hip rotations and knee flexion(ρc ≥ 0.7). Results were poor for hip adduction and abduc-tion and for ankle PROMs (ρc < 0.6).

In [see Additional file 4 Stable 4], we show mean measure-ment errors in degrees with 95% confidence intervalaccording to the graphs of Bland-Altman for all measure-ments. There is a tendency to overestimation for hip flex-ion and adduction. Hip abduction, hip internal rotation,knee PROMs and ankle dorsiflexion are generally under-estimated.

2. Study of reproducibility (intra-tester reliability)2.1. Pediatric orthopaedic surgeonBy using the confidence interval method of respectiveconcordance correlation coefficient, we demonstratedreproducibility for all PROMs.

2.2. ResidentWe demonstrated lack of reproducibility for hip adduc-tion and abduction while visual estimations of otherPROMs were found to be reliable.

3. Evaluation of the role of experience (inter-tester reliability)As visual estimations given by the resident are not reliablefor hip adduction and abduction, we deduce the impor-tance of level of experience for these PROMs. The role ofexperience was also demonstrated for hip rotations andankle plantarflexion using the confidence intervalapproach.

DiscussionThis study brings up the advantages and limitations of vis-ual estimation of PROMs in the pediatric lower limb.

This is to our knowledge the first study evaluating reliabil-ity of VE for hip PROMs. We found high level of accuracyof VE for hip flexion and rotations, and good accuracy forhip abduction, but a lack of reproducibility of measure-ments performed by a less experienced examiner. Con-cordance was poor for hip adduction (ρc ≈ 0.5). Manyfactors may have contributed to these poor results of VE ofhip adduction: the examination technique with the hip

flexed and the low absolute values of hip adduction com-pared to flexion, abduction and rotations. The level ofexperience was found to be important in estimating hipPROMs other than flexion.

Marks et al [26] investigated the reliability of VE of kneeROM taken by three physicians on patients with rheuma-toid arthritis, who examined the patients independently.One could assume that each examiner may have applied adifferent force to move the joint, theoretically leading tobias in interpretation of VE. Despite this possible bias, theauthors found good intra- and interexaminer reliability.In order to avoid a similar bias in our study, PROMs foreach joint were performed once by the same examiner.

Watkins et al [10] studied passive flexion and extension of50 knees. For each tested knee, two physical therapistswere chosen randomly from a list of 14 examiners. In thisstudy as well, estimations were done separately by eachexaminer in a position of his choice. Visual estimationswere performed before goniometric measurements foreach joint. We believe that this may influence subsequentvisual assessment by adjustment of visual estimations. Inour study, the two examiners who did visual estimationnever knew goniometric values of PROMs. Although theICC (Intraclass Correlation Coefficient used by Watkins etal [10] is less reliable than the CCC (Concordance Corre-lation Coefficient) [21] used in our study, we found simi-lar good concordance for knee flexion and highconcordance for popliteal angle. Therefore, our results aresimilar to those reported by previous studies [10,27]which found good accuracy for visual estimation of kneePROMs, with a tendency to slightly underestimate realvalues. Experience does not seem to play a major role inVE of knee PROMs. Even unexperienced examiners maysatisfactorily estimate knee motions without using a goni-ometer.

Youdas et al [7] evaluated active range of motion (AROM)of 45 ankles (dorsiflexion and plantar flexion). The studyincluded 10 physical therapists who performed measure-ments with their own preferences regarding position andmethod. They assessed interobserver reliability for visualestimation and intraobserver reliability for visual estima-tion and goniometry. AROMs are much more subject tovariations than PROMs. The authors found low concord-ance between VE of different examiners (ICC = 0,34 fordorsiflexion and ICC = 0,48 for plantarflexion) and a fairconcordance between visual estimation and goniometry(0,58 for dorsiflexion and 0,625 for plantarflexion). Webelieve that these low values are mainly due to the factthat AROMs are much more subject to variations thanPROMs and to the absence of standardization of gonio-metric measurements as recognized by the authors them-selves.



Allington et al [2] assessed intra- and interobserver relia-bility and reproducibility of goniometry and visual esti-mation of ankle PROMs in 24 children with spasticcerebral palsy (46 ankles). Two physical therapists per-formed all the measurements. They found very good cor-relations between the goniometer and the naked eye (r >0,94) for ankle dorsiflexion and plantarflexion. They alsofound a mean error (ME) of 5° with a SD of 5° in theinter- and intraobserver measurements. Thus, even withvery good reliability, there is a significant error to be takeninto account when performing visual estimates. In ourstudy, we found good level of concordance between thespecialist's VE and goniometry for ankle PROMs (ρc ≈0.7). We do not have an explanation for the very high con-cordance observed in the study of Allington et al [2] butwe believe that this may be partly due to the small numberof patients examined.

Disparity in results of VE of ankle PROMs between the dif-ferent studies is due to many factors: the number of exam-ined ankles, the maneuver technique and the method ofgoniometric measurement.

One of the major limitations of our study was the smallnumber of examiners. It is obvious that by increasing theirnumber with different levels of experience, more conclu-sions could be stated concerning the role of experience.Even though there were only two examiners performingvisual estimation, we believe that results can be extrapo-lated to other experienced and less experienced examin-ers.

Validity and reliability of goniometric measurements werenot verified in our study. This could have been done forexample by an additional examiner doing another set ofgoniometric measurements, but we really think that thisproblem was thoroughly discussed and verified in the lit-erature [1-16] and we've deliberately chosen goniometricmeasurements as a reference to study more precisely vis-ual estimation and the role of experience. Considering allmeasurements as variables to be verified would have cer-tainly complicated the statistical analysis and weakenedthe conclusions. Based on literature statements and onour standardized technique for goniometric measure-ments, we can assume with some caution that these meas-urements were valid and reliable.

While performing the examination, we sometimesencountered lack of compliance especially with youngerchildren, making measurements difficult to obtain. Inaddition, the force applied on the limb may vary for thesame range of motion not only from one child to another,but also for the same child when repeated measurementsare done. Different results could have been then obtainedif the same motion was repeated by each examiner. This

problem was discussed by Amis and Miller [28] whoexplained that passive motions are difficult to reproducebecause stretching of soft tissues at the limit of the motiondepends on the force applied on the limb. Wagner [29]found a greater variability when measuring passivelymotions of pronation and supination of the forearm.Kilgour (15) demonstrated that while some measurementerror arises during the placement and reading of the goni-ometer, the majority of it is most likely related to difficul-ties in determining end-range joint positioning. Perhapsforce dynamometers (including those that are hand-held)could be used to standardize the amount of passive forceapplied and thereby decrease the potential for error [30].We tried to limit the bias relative to these problems byexcluding children of three years or less, and by maintain-ing the same position while doing the three measure-ments (two visual estimations and one goniometricmeasurement). But while this may enhance reliability byreducing error, it does not reflect clinical practice whereexaminers perform ROMs separately. For this reason, wethink that a patient should be followed by the same ther-apist to document disease progression or to evaluate treat-ment results.

While comparing the charts of measurements (visual andgoniometric) of the two parts of our study, we noticed thatfor a given child, respective values of PROMs are differentand this difference is sometimes up to 20 degrees for bothmeasurements (visual and goniometric). This can behardly explained by the sole visual or goniometric errorand we are sure that a large part of variation is due tochange in PROM itself. To avoid such an important bias,we tested intra-rater reliability by comparing respectiveconcordance coefficients instead of direct comparison ofvisual estimations. We can also understand in this settingwhy it is misleading to study the reliability of goniometricmeasurements by comparing the two sets of measure-ments.

We had some difficulties with goniometric measure-ments. Short limbs were easier to measure because land-marks were closer to the goniometer's arms. This wasespecially true for motions around the hip and the knee.In retrospect, a longer armed goniometer would havedecreased some of the problems. On the other hand,some bony landmarks were moving under the skin. There-fore, the greater trochanter was marked while flexing thehip to minimize variations.

We had some overweight children. Obesity can make esti-mations difficult to obtain. It may also hide some bonylandmarks especially the greater trochanter. This is addedto the fact that, naturally, identification of bony land-marks is more difficult in the lower limb [31]. The AAOS[32] and Rowe [27] have suggested that visual estimation



is more reliable than goniometric measurements whenbony landmarks are not easily seen or palpated. However,overweight children were not excluded from our study toavoid the bias of selection.

ConclusionVisual estimation (VE) of passive hip flexion, knee flexionand popliteal angle is highly accurate and reliable regard-less of the examiner's experience. VE of hip rotations andabduction is also highly accurate provided it is performedby an experienced examiner. Goniometric evaluation ispreferred for passive hip adduction especially if it is per-formed by less experienced examiners. Inexperienced test-ers should not estimate hip adduction and abduction fortheir measurements are not reliable for these PROMs.Goniometry has to be used in such situations. The ankleseems to be the joint with the least reliability for VE ofPROMs, and this should be kept in mind and taken intoconsideration when reporting treatment results in scien-tific papers.

Competing interestsThe authors declare that they have no competing interests.

Authors' contributionsIG conceived of the study. RR, IG and IK performed thestudy and participated in its design. SEH participated inthe design of the study and in the statistical analysis. KKand FD reviewed the paper and made corrections. Allauthors read and approved the final manuscript.

Additional material

AcknowledgementsNo other contributors.

References1. Glanzman AM, Swenson AE, Kim H: Intrarater range of motion

reliability in cerebral palsy: a comparison of assessmentmethods. Pediatr Phys Ther Winter 2008, 20(4):369-72.

2. Allington NJ, Leroy N, Doneux C: Ankle joint range of motionmeasurements in spastic cerebral palsy children: intraob-server and interobserver reliability and reproducibility ofgoniometry and visual estimation. J Pediatr Orthop B 2002,11(3):236-9.

3. McWhirk LB, Glanzman AM: Within-session inter-rater realia-bility of goniometric measures in patients withspastic cere-bral palsy. Pediatr Phys Ther Winter 2006, 18(4):262-5.

4. Ten Berge SR, Halbertsma JP, Maathuis PG, Verheij NP, Dijkstra PU,Maathuis KG: Reliability of popliteal angle measurement: astudy in cerebral palsy patientsand healthy controls. J PediatrOrthop 2007, 27(6):648-52.

5. Mutlu A, Livanelioglu A, Gunel MK: Reliability of goniometricmeasurements in children with spastic cerebral palsy. MedSci Monit 2007, 13(7):CR323-9.

6. Menadue C, Raymond J, Kilbreath SL, Refshauge KM, Adams R: Reli-ability of two goniometric methods of measuring activeinversion and eversion range of motion at the ankle. BMCMusculoskelet Disord 2006, 28(7):60.

7. Youdas JW, Bogard CL, Suman VJ: Reliability of goniometricmeasurements and visual estimates of ankle joint activerange of motion obtained in a clinical setting. Arch Phys MedRehabil 1993, 74(10):1113-8.

8. Brosseau I, Balmer S, Tousignant M, O'Sullivan JP, Goudreault C,Goudreault M, Gringras S: Intra- and intertester reliability andcriterion validity of the parallelogram and universal goniom-eters for measuring maximum active knee flexion andextension of patients with knee restrictions. Arch Phys MedRehabil 2001, 82(3):396-402.

9. Brosseau I, Tousignant M, Budd J, Chartier N, Duciaume L, Plamon-don S, O'Sullivan JP, O'Donoghue S, Balmer S: Intratester andintertester reliability and criterion validity of the parallelo-

Additional file 1Correlation between specialist's VE and goniometric measurements. Abscissa axis represents the different PROMs; ordinate axis represents val-ues of the concordance correlation coefficient (ρc). HFX: hip flexion, HAD: hip adduction, HAB: hip abduction, HIR: hip internal rota-tion, HER: hip external rotation, KFX: knee flexion, PA: popliteal angle, ADF: ankle dorsiflexion, APF: ankle plantarflexion; 1(blue columns) and 2(red columns) correspond to the two parts of the study.Click here for file[http://www.biomedcentral.com/content/supplementary/1471-2474-10-126-S1.DOC]

Additional file 2Mean measurement error for all PROMs estimated by the specialist. Short horizontal lines represent mean error measurement in degrees; ver-tical lines represent confidence interval at 95%. HFX: hip flexion, HAD: hip adduction, HAB: hip abduction, HIR: hip internal rotation, HER: hip external rotation, KFX: knee flexion, PA: popliteal angle, ADF: ankle dorsiflexion, APF: ankle plantarflexion.Click here for file[http://www.biomedcentral.com/content/supplementary/1471-2474-10-126-S2.DOC]

Additional file 3Correlation between Resident's VE and Goniometric measurements. Abscissa axis represents the different PROMs; ordinate axis represents val-ues of the concordance correlation coefficient (ρc). HFX: hip flexion, HAD: hip adduction, HAB: hip abduction, HIR: hip internal rota-tion, HER: hip external rotation, KFX: knee flexion, PA: popliteal angle, ADF: ankle dorsiflexion, APF: ankle plantarflexion; 1(blue columns) and 2(red columns) correspond to the two parts of the study.Click here for file[http://www.biomedcentral.com/content/supplementary/1471-2474-10-126-S3.DOC]

Additional file 4Mean measurement error for all PROMs estimated by the resident. Short horizontal lines represent mean error measurement in degrees; ver-tical lines represent confidence interval at 95%. HFX: hip flexion, HAD: hip adduction, HAB: hip abduction, HIR: hip internal rotation, HER: hip external rotation, KFX: knee flexion, PA: popliteal angle, ADF: ankle dorsiflexion, APF: ankle plantarflexion.Click here for file[http://www.biomedcentral.com/content/supplementary/1471-2474-10-126-S4.DOC]


http://www.biomedcentral.com/content/supplementary/1471-2474-10-126-S1.DOC




















Publish with BioMed Central and every scientist can read your work free of charge

"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."

Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community

peer reviewed and published immediately upon acceptance

cited in PubMed and archived on PubMed Central

yours — you keep the copyright

Submit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.asp

BioMedcentral

gram and universal goniometers for active knee flexion inhealthy subjects. Physiother Res Int 1997, 2(3):150-66.

10. Watkins MA, Riddle DL, Lamb RL, Personius WJ: Reliability ofgoniometric measurements and visual estimates of kneerange of motion obtained in a clinical setting. Phys Ther 1991,71(2):90-6.

11. Ashton BB, Pickles B, Roll JW: Reliability of goniometric meas-urements of hip motion in spastic cerebral palsy. Dev MedChild Neurol 1978, 20:87-94.

12. Harris SR, Smith LH, Krukowski L: Goniometric reliability for achild with spastic quadriplegia. J Pediatr Orthop 1985, 5:348-351.

13. Stuberg WA, Fuchs RH, Miedaner JA: Reliability of goniometricmeasurements of children with cerebral palsy. Dev Med ChildNeurol 1988, 30:657-666.

14. McDowell BC, Hewitt V, Nurse A, Weston T, Baker R: The varia-bility of goniometric measurements in ambulatory childrenwith spastic cerebral palsy. Gait Posture 2000, 12:114-121.

15. Kilgour G, McNair P, Stott NS: Intrarater reliability of lowerlimb sagittal range-of-motion measures in children withspastic diplegia. Dev Med Child Neurol 2003, 45:391-399.

16. Gajdosik RI, Bohannon RW: Clinical Measurement of Range ofMotion. Review of Goniometry Emphasizing Reliability andValidity. Phys Ther 1987, 67(12):1867-72.

17. Ghanem I, Seringe R: Comparison of evaluation methods of theresults of congenital clubfoot treatment. Rev Chir Orthop Repar-atrice Appar Mot 1995, 81(7):615-21.

18. Fish DR, Wingate L: Sources of goniometric error at the elbow.Phys Ther 1985, 65(11):1666-70.

19. Palmer ML, Epler MF: Fundamentals of musculoskeletal assessment tech-niques 2nd edition. Philadelphia: Lippincott Williams & Wilkins; 1998.

20. Clarkson HM: Musculoskeletal assessment: joint range of motion andmanual muscle strength 2nd edition. Baltimore: Lippincott Williams &Wilkins; 2000.

21. Lin LI: A concordance correlation coefficient to evaluatereproducibility. Biometrics 1989, 45(1):255-68.

22. Barnhart HX, Haber M, Song J: Overall concordance correlationcoefficient for evaluating agreement among multiple observ-ers. Biometrics 2002, 58(4):1020-7.

23. Bland JM, Altman DG: Statistical methods for assessing agree-ment between two methods of clinical measurement. Lancet1986, 1(8476):908-9.

24. Bland JM, Altman DG: Comparing methods of measurement:why plotting difference against standard method is mislead-ing. Lancet 1995, 346:1085-87.

25. Lin LI: Assay validation using the concordance correlationcoefficient. Biometrics 1992, 48:599-604.

26. Marks JS, Palmer MK, Burke MJ, Smith P: Observer variation in theexamination of knee joints. Ann Rheum Dis 1978, 37(4):376-7.

27. Rowe CR: Joint measurement in disability evaluation. ClinOrthop Relat Res 1964, 32:43-53.

28. Amis AA, Miller JH: The elbow. Clin Rheum Dis 1982, 8(3):571-93.29. Wagner C: Determination of the rotary flexibility of the

elbow joint. Eur J Appl Physiol Occup Physiol 1977, 37(1):47-59.30. Bohannon RW, Lieber C: Cybex® II isokinetic dynamometer for

passive load application and measurement: Suggestion fromthe field. Phys Ther 1986, 66:1407.

31. Boone DC, Azen SP, Lin CM, Spence C, Baron C, Lee L: Reliabilityof goniometric measurements. Phys Ther 1978, 58(11):1355-90.

32. Joint motion: A Method of Measuring and Recording. Chi-cago, III: American Academy of Orthopaedic Surgeons; 1965:8.

Pre-publication historyThe pre-publication history for this paper can be accessedhere:

http://www.biomedcentral.com/1471-2474/10/126/prepub














































http://www.biomedcentral.com/1471-2474/10/126/prepub


http://www.biomedcentral.com/info/publishing_adv.asp


Documents

BMC Musculoskeletal Disorders BioMed Central · visual estimation of ROMs performed by residents [17]. By comparing values given by the POS to those of the res-ident, our purpose